Television Production Handbook 
Home Introduction Contents Ordering Information
1980-2009 Roger Inman & Greg Smith. All rights reserved.

Basic Video

Analog Video

Nothing could be simpler than picking up your camcorder and shooting a little videotape out in the back yard. At least on the surface. It's useful to remember that television itself was only a laboratory curiosity in 1940. In the 1950's you could make black and white television images in a studio, but you couldn't record them on videotape until 1956 when Ampex introduced the first commercial videotape recorder at the modest cost of $75,000 per machine. In the 1960's color television became common, and in the late 1970's videotape recorders found their way into businesses and schools. Cameras and recorders intended for individual use first hit the scene in the 1980's.

NTSC Video

Actually, there are three distinct analog television formats in use around the world today, with a number of variations. The American system (NTSC) was the first. The color signal was designed to work with the pre-existing black and white standard. The same broadcast bandwidth that was designed for high-quality monochrome signals was made to accommodate the far more demanding color signal. At the same time, color transmissions had to be compatible with existing monochrome receivers. Compromise was the word for the day, and quality suffered as a result. The PAL system was designed for color from the ground up, as was the SECAM system. PAL is used in England, Germany, and most countries having ties those two. The SECAM system is used in France, the Soviet Union, and countries associated with them.

There's no need to get into specific differences between standards. Since virtually all standard definition (4x3 aspect ratio) television in North America conforms to the NTSC standard, that's what we'll deal with.

The television picture is generated by a complicated combination of signals. It's useful to look at these signals, collectively called composite video, from two different perspectives. First, we'll look at the mechanics of displaying, or writing, the television picture in terms of time. Then we'll see how the parts of the video waveform are combined to make a picture appear on the screen.

Writing on the Screen

In non-electronic visual media (photography and film), each complete picture is displayed or seen in its entirety instantly. The television image, like a wire photo, must be "written" onto the screen dot by dot and line by line. This done in the United States according to standards set by the NTSC (National Television Systems Committee). Like film, television consists of a series of pictures. While film pictures are presented at a rate of twenty-four frames per second, television pictures are displayed at a rate of thirty frames per second.

The Television Scan

On television sets and monitors with picture tubes the picture is actually created by a stream of electrons striking a phosphor screen on the face of a CRT (Cathode Ray Tube). When a phosphor dot is hit by this electron beam, it glows for a fraction of a second. To aim the electron beam, two control systems are used. The vertical circuits move the electron beam from the top of the screen to the bottom, while the horizontal circuits move the beam from left to right. This movement from top to bottom and left to right is called the trace. The return path (bottom to top and right to left) is called the retrace. The electron beam is turned off during the retrace.
The NTSC picture consists of 525 horizontal lines displayed every thirtieth of a second. Instead of writing all 525 lines with each vertical trace, however, only every other line is written, odd numbered lines on one vertical trace and even numbered lines on the next. Each half of the picture is called a "field" and consists of 262 1/2 lines (half of 525). In order to write all 525 lines every thirtieth of a second, there must be two completed vertical traces for each picture. Thus, the field rate is sixty fields per second while the frame rate is half that, or thirty frames per second. A special series of signals, called equalizing pulses, is provided to make sure the two fields of video fit together, or interlace, properly.

Both PAL and SECAM signals have 625 horizontal lines, with a frame rate of 25 frames per second. Because of the slower frame rate both are able to display more picture detail than the NTSC system in the same bandwidth.

Composite Video

The portion of the waveform that causes the picture to be displayed on the screen in the right place at the correct rate is called "sync." The waveform that controls the brightness of the screen (by varying the strength of the electron beam) is called "video." Combined with the luminance signal is the signal that controls the hue and saturation of color in the picture. The three signals combined are called "composite video."

 Waveform monitor two horizontal lines
Two Horizontal Fields

Composite video is distributed through television systems through cable at a level of one volt peak to peak with an impedance of 75 ohms. The scale used to measure the amplitude of the video signal is divided into IRE (Institute of Radio Engineers) units. The picture part of the signal (disregarding the chroma portion) should lie in the area above zero up to 100 IRE units. The sync portion of the signal should be a series of pulses going from zero IRE units down to -40 IRE units.


 Waveform Monitor vertical interval

The Vertical Interval, with Equalizing Pulses

Sync consists of horizontal pulses, vertical pulses, and equalizing pulses. The total amplitude of the signal, from the bottom tip of the sync pulses to the brightest part of the picture, is 140 IRE units. At a standard impedance of 75 ohms, the correct amplitude of the composite video signal is one volt, peak to peak, so one IRE unit is equal to 1/140 volt peak to peak. Picture sources are set up with "black" at about 7.5 to 10 IRE and maximum brightness for all but brief transient spikes just below 100 IRE units.

Waveform Monitor Horizontal Interval  

The Horizontal Interval, with Burst Flag

A third signal, neither sync nor video, is always present in color pictures just after each horizontal sync pulse. This signal, called "burst," is a short sample of the color subcarrier frequency and is used as a reference to control the colors displayed on the screen. The burst signal goes from +20 to -20 IRE units. Chroma is not considered in setting black level (pedestal) and level (gain). However, if the video signal including chroma goes below -20IRE, the picture will interfere with the sync signal, which controls how the picture is displayed.

Before the advent of color television, the vertical and horizontal rates for NTSC television were actually 60 Hz and 15,750 Hz, respectively. In color television the rates had to be altered slightly to 59.94 Hz and 15,734 Hz, respectively, to prevent interference with the color subcarrier, which has a frequency of 3.58 MHz.

Component Video

The color NTSC signal is a clever interweaving of different elements. The color portion in broadcast cameras is actually generated by three pickup tubes or chips. The colors are red, green, and blue. There are some pieces of equipment that actually use four independent signals; red, green, blue, and sync. Some combine sync with one of the color signals.
These systems are simply called "RGB systems."

It's also common to combine the three color signals into two, called color difference signals. If fact, the color part of the broadcast television signal is a pair of color difference sidebands.

One of the most common "component" strategies comes from the way the signal is recorded on most VCR's. The color and luminance parts of the signal are recorded separately in a scheme called "color under." Every time the composite video signal is divided into luminance and chroma, and every time it's put back together into a composite signal, there is inevitable loss and distortion. Some systems can keep the luminance and chroma signals separate. The Sony Type V "RF Dub" edit mode is one example. "S Video" is another.

In the end, "component video" is not one standardized system, but rather any scheme that keeps parts of the total video signal separate in order to maintain higher quality. The problem is that these various formats are not directly compatible and the only universal NTSC standard for moving pictures from one machine to another is composite video.

Writing the television signal on videotape

Almost all videotape recorders on the market today are "helical" recorders. All machines record from one to four audio signals on tape just as they would be on an audiotape recorder. Some are recorded by fixed heads in a linear fashion along the length of the tape, while others are recorded by the video heads.

The video signal contains enormous amounts of information. Just as the amount of detail in a photograph is dependant on the fineness of the grain structure of the film, the amount of information you can record on tape is dependant on the oxide structure on the tape. To record enough information for a full broadcast television signal using a fixed video head, the tape would have to be moving at over five hundred inches per second. Instead, two (or more) video heads are mounted on a drum rotating at 1800 rpm. Each head writes a diagonal stripe on the tape representing one field of video. A pair of heads, working together writes two stripes, representing one full frame of video. As the tape moves over the rotating video head drum, a series of diagonal stripes are written onto the tape.

Finding these stripes of video is the job of the last recorded signal, the control track. This signal is nothing more than a 60 Hz pulse recorded on the edge of the tape. Since the physical relationship between the diagonal video tracks and the individual control track pulses is fixed during recording, it's possible to use these pulses to find the video tracks on playback. The process of finding and reading the video tracks during playback is called "tracking." Failure to find the video tracks causes "snow" or noise in the video, or even complete picture loss.

Compromises

To record the full broadcast spectrum on tape, the original tape format was two inches wide and moved laterally at 15 inches per second. The Ampex videotape recorder had an effective head speed of 1500 inches of tape per second.   Making the tape smaller and slowing it down required three key compromises.

The first was a reduction in signal-to-noise ratio. Even tapes recorded on one inch VTR's can be rerecorded up to five or six generations before there's noticeable noise in the picture. On the other hand, VHS tapes have noticeable noise in the first generation and objectionable noise by the third.

The second compromise was in resolution, or frequency response. The luminance bandwidth in broadcast television is 4.2 Megahertz, with a horizontal resolution of 336 lines. The luminance bandwidth of VHS tape is about 3 Megahertz, allowing a resolution of 230 to 240 horizontal lines.

The third compromise involves using the "color under" technique to record a very low resolution version of the color signal at a frequency below the luminance signal, on the assumption that a relatively sharper monochrome signal will mask the fuzzy color signal. It works, to a point. In fact, first generation VHS or 8mm tape looks reasonable.

You can extend the resolution of a recorder if you're willing to accept more noise. Noise becomes objectionable at about three IRE units, or an overall signal to noise ratio of about 35dB. You can fake resolution by using "image enhancement," which is a process that amplifies transitions between light and dark parts of the picture. Too much "enhancement" gives the overall picture a pasty look, especially in areas of fine detail.

While these tricks might make a first generation recording look somewhat better, as you copy a tape from one generation to the next the effects of all of your compromises and tricks are compounded. In the end, you can't beat the laws of physics. In the analog world, there are really only three ways to get better pictures on tape. The first is the traditional solution of moving the head over more tape by making the tape wider or by increasing the linear tape speed. The second is to improve the tape by reducing the size of the magnetic particles. This is the route taken by moving from metal oxide tapes to metal tapes. The third is to keep the luminance and chroma signals separate in some sort of "Y/C" scheme.

Digital Video

The first digital video format, D1, was introduced in 1986.  Unfortunately there were no personal computers with enough storage, memory, or processor speed to use it. Within four years Newtek had a viable add-on for the Amiga computer that could function as a video switcher, graphics generator, and 3D modeler.  Still, there was no practical way to capture or store digital video.  By the mid 1990’s we were using video capture cards to convert the NTSC video signal to digital computer files and using nonlinear editing software.  Digital video has no need for synchronizing signals, so only the picture content of the NTSC signal was digitized as a 640 x 480 pixel frame, the same as a VGA computer image, and maintaining the 1.33 aspect ratio.  

By the year 2000 the DV (digital video) standard allowed direct digital recording on videotape.  Unlike the NTSC standard, the digital video recording consists of 24 bit color images, 720 pixels wide by 480 pixels high, that can be sent directly to disk as a computer file.   DV is also referred to as SD, or Standard Definition, as opposed to HD, or High Definition.  SD digital video does not have square pixels.  To improve resolution over a standard 640x480 pixel VGA display, digital video pixels are nine tenths as wide as they are high, allowing for 720 horizontal pixels by 480 vertical pixels with the same 1.33 aspect ratio.

 There is a second version of SD digital video which has pixels that are 1.2 times as wide as they are high.  The same 720 x 480 picture in this format yields an aspect ratio of 1.78, or 16 x 9.  It is not advisable to shoot in 16x9 SD unless you are sure that your software and hardware will recognize this format and play it back correctly at each stage of post production, including nonlinear editors, conversion programs for MPEG2, Windows Media Files, Real Player, or Quicktime, DVD authoring software, and playback equipment.  Compression schemes that look good with .9 to 1 pixels may not work as well with 1.2 to 1 pixels, especially where fast cameral movement is involved.  Before shooting 16x9 SD, test every link in your production chain.

The 24 bit color scheme allows for eight bits each to represent red, green, and blue.  In effect, digital video is component video in digital form.  Each eight bit color can have 256 distinct steps from darkest to brightest.  The total number of possible colors is 256 cubed, or 16,777,216.  Any device or system that has “16 million colors” is 24 bit color.

Unlike the original D1 digital standard, the DV standard is compressed to about one fifth the size of an uncompressed file.  The data rate is about four megabytes per second.  The way in which the data is compressed can vary from one manufacturer to the next.  Even though audio/video files for Windows generally have the .avi extension (.mov for Apple), the ability to read (or decompress) these files resides in the hardware or software used for the original compression.  The method of compression and subsequent decompression is called a codec.  Literally  dozens of codecs have been used by various capture card and video editing companies.  Microsoft supplies most of the popular codecs, while others can be downloaded from sites on the internet.  Still, always keep the original raw videotape if you can.

Since the move to digital broadcasting, we are using a new collection of standards.  While NTSC stands for the National Television Standards Committee, ATSC stands for Advanced Television Standards Committee.  There are many different ATSC standards, all dealing with digital broadcasting.  The three basic video formats are SD (1.33 aspect ratio with 720 x 480 non-square pixels and HD, either 720P or 1080i.  720P has 1280 x 720 pixels, while 1080i has 1080x 1920 pixels.  HD pixels are square.  The aspect ratio is 1.78, or 16 x 9.

Interlace and Progressive Scan

One of the fundamental aspects of NTSC television is becoming an historical curiousity.  To avoid the flicker of 24 fps film, video was designed to write two fields every thirtieth of a second, the first field containing the odd numbered horizontal lines and the second field containing the even numbered horizontal lines.  The MPEG standard for broadcast television no longer requires this interlaced image.  720p consists of writes all lines for each frame every thirtieth of a second for broadcast.

Video Sources

Although there are many sources of video, all fall into two categories: electronically generated sources and optical sources. Electronic sources include all of the standard test signals, character generators, computers, and background generators. Optically generated video is produced by television cameras, either directly or prerecorded on videotape.

The electronic test signals are used primarily to set up monitoring equipment and are rarely included in program content. They include color bars, gray scale, dots, crosshatch, and multiburst.

Color Bars

Color Bars

While color bars have several formats, all share a sequence of colors and gray scale values that are fixed and can be used as a standard reference in adjusting equipment. The color sequence is always white, yellow, cyan, green, magenta, red, blue, and black, reading left to right. Each color, going from left to right, has ten per cent less luminance than the preceding color. Using this signal it's possible to set a monitor for the correct contrast and brightness, as well as for the correct color saturation and hue, guaranteeing an optimum picture display. Color bars are often used at the beginning of a program to allow setup of tape processing equipment and playback receivers and monitors.


 Gray Scale

Gray Scale
 
Gray scale is either a five or a ten step scale plus black used to set up monitors for proper brightness and contrast range.


 Crosshatch

Dots and Crosshatch

Dots is a matrix of white dots on a black background. Crosshatch may be shown as a series of vertical or horizontal lines or as a combination of both. Dots and crosshatch are used to adjust the electron beams in color monitors so that the red, green, and blue signals overlap properly. Neither signal has any burst or chroma content. Crosshatch can also be used to make sure graphics shown by cameras are straight.


Multiburst

Multiburst

Multiburst is a series of square waves with increasing frequency from left to right. It's used to check the overall frequency response of various pieces of video equipment and is used primarily as an engineering tool.
The remaining electronically generated signals are used in the actual production of programs.

Color Black

The most basic is called "color black." This signal has a luminance value of 7.5 IRE and no chroma in the picture. It does have sync and color burst. It's the signal used every time the picture "fades to black."
 Color Background
The color background generator provides a solid color background in which the luminance, chroma, and hue can be adjusted. It's normally used as a background for text or graphics displayed on the screen. It's possible to adjust the background generator to create color fields that exceed the technical specifications for video. This results in improper operation of home receivers and should be avoided if at all possible.

Character Generator

The character generator is used to write and display text on the screen. It's actually a microcomputer with the ability to store and display text in a video format compatible with NTSC color signals. Some models provide for the ability to move text horizontally across the bottom of the screen or to roll text from bottom to top. Many character generators also may have the ability to store material on magnetic disks.

Computer

The computer is the tool of choice for artists working in television and graphics. The ability to capture and digitize live or videotaped images, manipulate them, and store them for later use is essential to every television producer. Most computers are capable of capturing and editing video. They can be made to output NTSC-compatible video with the addition of a video card that supports a video output.
 
Artwork done for SD (Standard Definition) television should have an aspect ratio of 1.33.  In graphics systems using square pixels the image should be 640 x 480 pixels.  In systems using .9 x 1 pixels, the image should be 720 x 480 pixels.

Film Chain

The film chain, for broadcast purposes, is obsolete.  A specialized camera is incorporated into the "film chain," a device used to show 35mm slides and motion picture film. This camera is locked in a fixed location. Often it looks at a device called a "multiplexer," which sends one of three optical picture sources to the camera. A mirror is positioned by remote control to select the picture source. Slide projectors normally have two drums or carousels and can "dissolve" from one slide to the next. The film projectors have special shutter systems to adjust the normal film speed of 24 frames per second to the television frame rate of 30 frames per second.

Analog Videotape

Video recorded on tape can be played back through a special computer called a "time base corrector." The signal from a videotape is not very stable compared to live sources, nor is it "locked" to the signal generator which drives all live sources. The time base corrector has the ability to take the picture from a videotape recorder, store it for a very short time, and send it on after adjusting the timing of the signal to conform to other video sources. Time base correctors usually have controls to adjust pedestal (black level), amplitude (gain), chroma amplitude, and chroma shift for color correction, in addition to controls to shift horizontal and vertical timing and the width of the horizontal interval. Tapes played through a time base corrector do, however, have to conform to reasonable technical standards and must have proper sync signals. Many monochrome cameras are incompatible with time base correctors because they lack essential sync signals or process sync in a way that is incompatible with the NTSC color system used in production studios for American television. The most common problems are lack of 2:1 interlace of video fields and lack of equalizing pulses.

Digital Video

Digital videotape comes in a number of sizes and formats.  All can be copied directly to computers via USB or IEEE1394 "Firewire" ports with appropriate software.  Some video cameras use DVD-RAM disks that can be played by computer DVD drives.  Others have their own hard drives that can be attached to computers as external hard drives.  Finally, a number of  cameras can record directly to removable memory cards that can be read by computers with card readers.

All of these recording systems avoid the timing problems inherent in analog recordings, but digital tape still can suffer from dropout caused by oxide problems.

Computer Disks

Both CD’s and DVD’s can be used as data disks with video content.  Both can be used in live environments as video sources when connected to a computer with the proper video output.  DVD players can be used to play back DVD’s or (in some cases) CD’s.  It is hard to imagine a situation, however, in which it would not be better to introduce the content from a CD or DVD during editing, rather than during the recording of a live event.

Synchronization

Because of the complexity and the technical requirements of the video signal, pictures from different sources can't be combined without first making sure that all sources are synchronized. The biggest challenge in wiring a television control room or edit suite is to make sure that each signal reaching the special effects generator begins a new frame of video at the same time as every other incoming signal. To accomplish this, every color television facility uses a signal generator which provides timing signals to all video equipment which must interact. In addition, all signals must arrive with the color reference, or burst, at the same phase angle as all other signals.

Some cameras and recorders can't be locked to an external sync source. Those that can use either video or three discrete synchronizing signals. Those that use video have a "genloc" input which is used to lock an internal sync generator to the incoming signal. Those requiring discrete synchronizing signals use vertical drive, horizontal drive, subcarrier, and perhaps composite sync, produced by a special signal generator.

Timing and phase errors between sources are potential causes of bad pictures. If a picture problem is unique to a dissolve, wipe, or key situation, timing and color phase are possible causes.

There are a number of issues related to signal timing that keep television engineers busy. Just knowing that timing is important any time you connect or mix the outputs of two video devices should be enough to alert you to the cause of many otherwise mysterious problems.

Monitoring

Waveform Monitor

 Waveform Monitor

In judging pedestal and gain, and technical signal quality, a special oscilloscope called a waveform monitor is used. This monitor has a number of operating modes, but the most commonly used gives a display of two horizontal lines of picture, showing the combined effect of both the luminance and chroma portions of the picture. It's calibrated in IRE units, from -40 IRE to 100, so precise measurement of the amplitude of both sync and picture is possible. It's also possible to isolate either the luminance portion or the chroma portion of the picture and to look at the horizontal sync portion alone, the vertical sync portion alone, or at two fields of video. Waveform monitors are valuable tools in evaluating the timing and technical quality of video sources and are therefore useful as trouble-shooting tools as well as production aids.

As you know, the amplitude of the video signal is set at one volt peak-to-peak into an impedance of 75 ohms. A glance at the waveform monitor can reveal one of the most common technical problems in television. Many devices allow you to "loop" an input through the device to another. Picture monitors generally allow you to do this. They also have a termination switch next to the looped input. Along the line from your picture source only one device should be terminated. A second termination will cut the signal amplitude in half. Lack of any termination will double the amplitude. If the sync portion of the composite signal on the waveform monitor is either much larger or much smaller than -40 IRE, the problem is almost always improper termination.

Vectorscope

Vectorscope
 
The vectorscope is used in analyzing the color content of the video signal. It's set up like a polar graph, showing hue as a phase angle with reference to the "burst" signal, which should always be at zero degrees. Amplitude (distance from the center) of a color indicates the saturation of the color. One of the main functions of the vectorscope is the adjustment of color phase of all color video sources in the system. Using a standard reference signal, usually the color subcarrier or color black, each color source is examined and the phase angle of its "burst" signal adjusted to zero degrees with reference to the standard signal. Only if all sources are "in phase" can they be used in special effects such as wipes and dissolves without objectionable color shifts. The vectorscope is also useful in evaluating the accuracy of color rendition of cameras and in setting up color background fields.

Meters

Finally, videotape recorders and some other devices have video meters. These are not nearly as precise as the waveform monitor, but can be used to confirm the presence of a video signal and whether that signal is in a reasonable range. Although these meters may not have any calibration scale indicated, they do have a colored section which conforms to a reasonable video level for most color signals. The most obvious exception to this is color black which gives a meter reading well below the "optimum" level. Video meters have only limited usefulness.

Switching

The heart of any video control room is the special effects generator, or video switcher. This is the device used to select pictures from the various video sources and to create the effects which are the "language" of television. While the average video switcher offers a bewildering number of possible effects, it's usually laid out in a logical and functional manner easily mastered if it's taken one step at a time.

The most common special effect is the "take." It's the instantaneous change from one picture source to another. The reason it's the most common is that it's almost always the most appropriate way to change picture content in the course of a program. It may seem strange to use such an imposing piece of equipment as a video switcher for such a simple task, but it's crucial for any director to remember that all of the equipment around him serves only one central purpose: to transmit information to his audience cleanly and without confusion or distraction. The program content must carry and maintain the interest of his audience. Any attempt to "dress up" a program with unnecessary effects will simply distract or confuse the audience and will, therefore, be self-defeating. There is a video "language," just as there is a film language and just as there are spoken and written languages. In all of these it's the ability to communicate clearly that is the mark of the literate person.

The next commonly used effect is the dissolve. This is a gradual cross-fade from one picture to another during which one picture is superimposed on the other. The speed of the dissolve should be determined by the pace and mood of the program material. The dissolve can be used to establish or enhance a lyrical mood. But it can also be used for major transitions in program content that need to be softened; that is, where a take would be too jarring. Often a dissolve is used to indicate the passage of time or a change of location in dramatic programs. In other contexts, it might be used in transition from a wide shot to a close-up to emphasize continuity.

The wipe has less ambiguous uses. When a complete wipe occurs, one picture is replaced by another as though it were pushing the other off of the screen. Most switchers offer a number of wipe patterns, including horizontal, vertical, various corners, circles, and perhaps some others of limited usefulness. The wipe is used in two ways. First, it's used to insert a portion of one picture into another, allowing the viewer to see two scenes at the same time. The wipe is sometimes used in the same way as the dissolve to indicate a change in time or location.

Keys and mattes are used to insert one picture into another. Luminance keys use the monochrome signal as a switching device. Any portion of the keyed signal reaching a specific video level replaces the video content of the background signal. The amount of luminance necessary to effect this replacement is adjusted to achieve the desired effect. A matte is similar to a key, except that a solid color from the background generator is inserted into the background video in place of the video from the key source. Chroma keys work on the same principle, but use the color saturation of a specific hue as the keying signal rather than the luminance level.

Luminance keys and mattes are used almost exclusively to superimpose text or graphics on the screen. A character generator, for example, uses a luminance key to put text over normal video. Chroma keys are used to insert more complicated video into the background. Because of the wide range of luminance values in video containing people or objects, a clean luminance key is often impossible. By carefully controlling the hues in a scene, though, it's often possible to generate a clean chroma key. The most common color used in chroma keys is deep blue. The most common use of chroma keys is the insertion of video tape or computer graphics into newscasts.

Handbook Contents